15:01:05 #startmeeting Fedora ARM and AArch64 Status Meeting 15:01:05 Meeting started Tue Apr 24 15:01:05 2018 UTC. The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:05 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:01:05 The meeting name has been set to 'fedora_arm_and_aarch64_status_meeting' 15:01:05 #chair pwhalen pbrobinson 15:01:05 #topic roll call 15:01:05 Current chairs: pbrobinson pwhalen 15:01:19 morning folks, who's here today? 15:01:42 * pbrobinson o/ 15:03:24 lets give it a minute for others to join 15:05:12 howdy jlinton_ 15:06:35 ok, lets get started.. 15:06:41 #topic 1) ==== Userspace Status ==== 15:06:52 any new userspace issues? 15:07:27 has anyone has time to look at the firefox issue? 15:07:35 #info [abrt] firefox: raise(): firefox killed by SIGSEGV 15:07:35 #link https://bugzilla.redhat.com/show_bug.cgi?id=1564204 15:08:09 while not a blocker, its quite unfortunate for those that run a desktop 15:10:07 #info Help with Firefox debugging is appreciated. 15:11:06 it can also be fixed with a zero day update and I noticed that the first of Firefox 60 builds started to land so lets wait to see how that looks 15:11:40 ok, any other userspace issues? 15:12:05 pbrobinson: What happened with zlib? 15:12:23 I didn't see the package updated, (did I miss it), or is that patch just to ugly... 15:12:36 all 3k lines of it. 15:13:05 jlinton_: I had issues with build and I've not got back to it yet, going to try to this week 15:13:36 Without 2/2 or with 2/2? 15:13:56 I seem to be the only one that saw perf regressions with the last patch 15:15:17 jlinton_: don't remember sorry, will ping with issues later in the week 15:15:18 * jonmasters_ on 15:16:04 ok, moving on.. 15:16:08 I've wondered if converting the .spec to cmake is the right idea 15:16:15 or not 15:18:27 #topic 2) ==== Kernel Status ==== 15:18:57 any new kernel issues to report? 15:19:08 jlinton_: you mentioned an issue with hikey, I don't remember the details though 15:20:40 ? I haven't spun up 4.17 on the hikey yet, but i'm not away of any "new" issues, outside the normal bumps. 15:21:15 s/away/aware 15:21:29 jlinton_: something in dracut maybe? We discussed it at Connect I don't remember whether we resolved it 15:22:22 Yes, thats been a problem, the reset driver isn't in the initrd by default so the mmc doesn't work as a boot devbice 15:24:02 jlinton_: OK, thanks 15:24:06 I haven't reverified the juno either 15:25:06 (its actually the SD, that doesn't work on the hikey, not the emmc which I disabled in DT to get a working efi variable store, but the same problem probably happens with the emmc if a dt is used with that enabled) 15:26:17 yep 15:26:23 anyone else with HW issues? 15:27:04 (the driver, seems to be hi6220_reset.ko which is missing on the hikey) 15:28:02 #info Bug 1531140 - no console output with acpi on mustang (Warning: unable to open an initial console.) 15:28:26 #info Fixed in kernel-4.16.3-301.fc28 15:29:15 any other kernel news? 15:29:25 jonmasters_: anything on kernel? 15:31:07 #topic 3) ==== Bootloader Status ==== 15:31:28 nothing particularly major around bootloader 15:32:01 #info No issues reported. 15:32:16 #topic 4) == F28 Testing == 15:32:47 #link https://fedoraproject.org/wiki/Test_Results:Fedora_28_Branched_20180421.n.1_Summary 15:33:00 thats the latest nominated nightly 15:33:09 I did a bunch of testing over the weekend. Overall I think things are looking OK 15:33:46 #info Blocker Bugs: https://qa.fedoraproject.org/blockerbugs/milestone/28/final/buglist 15:33:52 #info please test Fedora 28, we're _VERY_ close to GA so now is the time to test and report issues you see, reply to the meeting notes to ensure they're picked up 15:34:15 I plan on reinstalling most of my victims later this week after I repost the PPTT stuff 15:35:47 hrm, looks like openqa hasnt tested the latest nightly. will inquire 15:36:05 #topic 5) == Open Floor == 15:36:33 so with the sun setting on F-28 dev people should be thinking about what's needed and wanted for F-29 15:36:45 If anyone is testing F28, and wants to see if the zlib patches are faster there is a koji build here: 15:36:58 https://koji.fedoraproject.org/koji/taskinfo?taskID=26292536 15:38:00 I think it might be worth the effort to broaden the openqa tests, and start running through all the base packages and assuring that `make check` etc is enabled 15:38:43 to many of these bugs rely on a person installing the system and noticing that firefox/hexchat/whatever doesn't actually work 15:39:14 jlinton_, we mirror whats tested on primary, but agree expanding those tests would be great. 15:39:37 right now its a little resource constrained runnign the tests we do run, currently running on a mustang 15:40:20 fwiw, firefox was caught in openqa, it tests cockpit and was failing 15:41:33 Oh nice.. If its a question of a couple fairly low end machines (mcbins/whatever) i'm sure I can find the right people to provide them. 15:41:38 but with that acpi issue for xgene fixed, we're planning on moving openqa to the m400 15:42:19 (thats the ongoing issue with the console that isn't upstream?) 15:42:53 jlinton_: well shitty firmware, msalter proposed a quirk 15:43:20 jlinton_: https://patchwork.kernel.org/patch/10351797/ 15:43:24 Yah, I'm assuming its the resource producer/consumer problem I was involved in the RH one a release or two ago, 15:44:59 That isn't the same patch, but it sounds like another fix for the same basic problem. 15:49:27 sorry folks, distracted. anything else for today? 15:50:10 #endmeeting